The present invention relates to a semiconductor device, and more particularly, to a semiconductor device including a multiple-layer hard mask with opposite stresses between layers and a method for fabricating the same.
Various processes have been performed to reduce undesirable events generated by decreased process margins between layers when forming a semiconductor device. For instance, gate lines and bit lines are using a multiple-layer structure to increase a self-aligned contact (SAC) margin during contact hole definition.
One of the limitations related to the usage of the multiple-layer structure is a peeling event. The peeling event refers to a portion of a layer in a stack structure peeling off due to differences in stress between layers and an increased stress change between the layers due to various subsequent thermal processes.
FIG. 1 illustrates a diagram showing a structure of a typical pattern. A first hard mask 12 is formed over a tungsten (W) layer 11. A second hard mask 13 is formed over the first hard mask 12. The tungsten layer 11 functions as a bit line and is denoted as ‘BL W’. The first hard mask 12 may include a nitride-based layer and is denoted as ‘HM NIT’. The second hard mask 13 may include a tungsten layer and is denoted as ‘HM W’.
An initial stress after formation for the tungsten layer 11 and the second hard mask 13 is a tensile stress (TEN), and for the first hard mask 12 is a compressive stress (COMP). When a nitride-based layer and a tungsten-based layer contact, each layer typically obtains an opposite stress to each other.
The initial stress changes by the heat from a subsequent thermal process 100 and thus a final stress having a different property than the initial stress results. For instance, the tungsten layer 11 and the second hard mask 13 may obtain a compressive stress and the first hard mask 12 may obtain a weak tensile stress, as represented with ‘TEN - - - ’, by a subsequent process for forming bit line spacers 14 accompanied with heat. At this time, the second hard mask 13 may obtain a strong compressive stress as represented with ‘COMP—’. Accordingly, the initial stress and the final stress obtain an opposite property to each other by the subsequent thermal process. Reference numerals S1 and S11 refer to an initial stress difference S1 and a final stress difference S11 between the first hard mask 12 and the second hard mask 13.
However, in the typical pattern, subsequent processes may generate undesirable contacting events between the first hard mask 12 having the weak tensile stress (TEN - - - ) and the second hard mask 13 having the strong compressive stress (COMP—) formed at the top, causing a peeling event. That is, an interface becomes weak as the difference in stress becomes large between the first hard mask 12 and the second hard mask 13. Thus, the second hard mask 13 may be easily seceded by subsequent etch processes and cleaning processes.